Can AI ever truly capture the emotional connection of live music?

In seconds, AI can generate a knockout pop song or a pastiche of a Mozart symphony with a single text prompt. It won’t play out of tune or run out of breath and it always plays the right notes. There’s no diva making demands, no grumpy band on the road or tortured creative.

But is it really music? As a conductor, I work with the raw materials of music. My job is about nuance and interpretation – we might ask the violins to lean into a warmer vibrato, or the brass to give more punch to a rhythm. We’ll negotiate tempo and try different phrasing in rehearsal, turning the notes on the page into something meaningful. What does the composer want to say?

Yet music goes beyond artistry and storytelling – it is intrinsic to being human and has deep biological effects. Our hearts “dance” to it in real time, spiking or lowering our pulse; an hour of drumming can boost immunity and a tempo between 60 and 80bpm can help us drift off to sleep faster. At the edge of life, the power of music to reduce pain and ease breathing in palliative and end-of-life care is profound.

AI music is everywhere – on your Spotify playlists, TV adverts, in your gym – and some of it is good. It can generate perfect, blemish-free music, but can it really build an emotional narrative? Recent research shows that while we may struggle to identify the difference between human- and AI-generated pop music, we consistently rate the handcrafted music as more emotionally intense.

While it’s cheap, quick and easy, AI poses an existential threat to the live music ecosystem. Many musicians make their living playing for film and TV scores, video games, adverts and earning streaming royalties. Grassroots venues cannot compete with the (highly entertaining) mega-budget avatar shows like Abba Voyage. If AI developers continue to scrape catalogues of copyrighted music, paying no royalties or licensing fees to the original artists, human musicians face extinction.

A live performance moves us on a fundamental level. I know first-hand the power and emotional weight of an elongated silence, an overwhelmingly long crescendo with 100 musicians giving their all, a devastatingly sorrowful singer… compounded by the danger of it all going wrong.

Live music stimulates endorphins and euphoria and reduces feelings of loneliness. And what still blows my mind: whether we know each other or not, our heart and respiration rates synchronise together in a concert hall.

AI music is everywhere – on your Spotify playlists, TV adverts, in your gym...

This year I’ve been incredibly lucky and privileged to sit as a series judge on BBC Young Musician, one of classical music’s biggest competitions and prizes. I’ve been struck by the imagination, creativity and resilience of our young competitors. The advocacy they show for music and live performance is infectious – they spend hours perfecting their technique and musical skills before stepping before us and unleashing their incredible music-making. There’s no space for AI-generated music in their world.

But there’s no escaping the growing presence of AI in our lives, and there are benefits like playlist curation and new music discovery. But to my mind, it is impossible to re-create and capture the human qualities that make music and live performance special.

I want to experience music in the danger zone, infused with life experiences and emotions. I want to feel the goosebumps and the synchronisation with those around me. I want to feel
the hard work and physical challenge of performance. That’s human connection. That’s what music really is.